AI, governed.
Data, sovereign.
Generative AI is already inside your organisation — in browsers, inboxes and side channels. We make it productive, compliant and Swiss: policy first, technical enforcement second, sovereignty always.
The AI redaction gateway.
Policies alone don't stop a paste into a chatbot. Our gateway sits between your users and any LLM, detecting and pseudonymising personal data, client names and secrets in real time — before anything leaves your perimeter. Names come back where they belong: inside it.
Three ways in, one coherent programme.
Govern
AI usage policy, risk register, tool inventory and staff training — aligned with ISO 27001, nLPD and the EU AI Act, and written so people actually follow them.
- AI policy drafting & rollout
- Risk assessment & registers
- Awareness training
Protect
Deploy the redaction gateway across your LLM traffic, with detection tuned to your own entities — clients, projects, personnel — not just generic patterns.
- Gateway deployment & tuning
- Custom entity recognition
- Audit logging & reporting
Deploy
For the most sensitive workloads: language models hosted in Swiss datacenters or on your own hardware. Full capability, zero data export.
- Swiss-hosted LLM APIs
- On-premises model deployment
- Use-case engineering & integration
Your data is your mandate.
For an NGO, leaked beneficiary data is a safeguarding failure. For a fiduciary, leaked client data is a professional-secrecy breach. Ungoverned AI turns everyday productivity into an export channel — unless the controls are in place before the habit is.
Let's look at your environment.
A one-hour assessment with a senior engineer. No sales script, no obligation.